Data Customer

AAKANKSHA ASHOK JHA

Tempe, Arizona ****** 480-***-**** https://www.linkedin.com/in/aakankshajha/ ******@***.*** Summary:

Data professional with 5 years + of industry experience in Consulting and Machine Learning with background in Statistics and Computer Science. Education:

Master of Science in Business Analytics (MSBA), W. P. Carey School of Business at Arizona State University, Tempe, AZ May 2018 Bachelor of Science in Computer Science, Mumbai University, Mumbai, India June 2014 TECHNICAL SKILLS

• Language: R, Python (pandas, numpy, scikit- learn, seaborn), SQL

• Data Visualization: Tableau, R ggplot2, Matplotlib, IBM Analytics

• Big Data: Hadoop, Pig, Hive, Power BI, Kafka, AWS

• Marketing Tools: A/B Testing, Market Mix Model

• SAP: SD, TSW, LE, CT, FI, BI, APO & Service Now

• Other tools: SPSS, Microsoft Azure ML, Minitab, VBA Excel, SAS

• Quality Management: Lean Six Sigma & Design of Experiments

• Enterprise Analytics

• Data Modeling & Optimization

• Database Management

• Recommendation Engines

• Text Mining & Sentiment Analysis

• Social Network Analysis

• Clustering & Classification

• Statistical Analysis & Inference

• Hypothesis Testing,

Confidence Intervals, ANOVA

• Regression Modelling, Neural Networks

• Feature Selection / Engineering

• Time Series Forecasting

• Naïve Bayes & Multivariate Statistics

WORK EXPERIENCE (5 Years +)

APL Logistics, Arizona, United States

Data Scientist Jan 2018 – May2018

• Implemented an ensemble of various Predictive Analytics algorithm like Support Vector Regression, Boosted Decision Tree, Neural Networks to predict Wins/Loss of bids to focus on important customers. Developed a 5-fold cross validation with hyperparameter tuning for increasing prediction accuracy.

• This helped in customer Retention with an accuracy up to 87% which substantially increases the winning bid percent by 1.5 times.

• Identified top features using PCA and combined it with Recency Frequency and Monetary Parameters from RFM Analysis.

• Performed RFM analysis for Customer bucketing and created a Tableau dashboard for visualization RFM and storytelling to business stakeholders.

• Managed a team of 4 and functioned as a point of contact for the team and the supervisor. Accenture, Mumbai, India

Consultant Jan 2016 – July 2017

• Predicted demand for a retail client based on historical data using timeseries forecasting model like ARIMA to meet seasonal customer requirements.

• Predicted Customer Churn using classification algorithms like Random Forest, Decision Jungle, Boosted Decision Tree with an accuracy up to 81.2%

• Classified the potential churning clients as loyal/risky/potential/lost and devised marketing strategies to control churning for a Fortune 150 client.

• Implemented Sentiment Analysis for customers’ comments & behavior metrics before & after deployment to know whether to continue with change

• Enhanced systems, decision making capabilities for a fortune 100 client by leading a team of 9 members.

• Analyzed large data sets and gave business insight using Business Intelligence tools and help clients block the future or past customers who would default.

• Collaborated with teams to optimize on employee scheduling based on idle time per employee in project which helped in cost cutting up to30%. Associate Consultant Aug 2014 – Dec 2015

• Utilized customer credit history and perform customer binning and give them credit limit based on their background and past transactional behavior.

• Predicted Customer Default Rate for Credit Cards using Logistic Regression with Stochastic Gradient Descent on 3 factors with ROC of 77%.

• Analyzed and fetched sensitive data by applying advance SQL Queries for U.S Bank Database.

• Performed Market Basket Analysis for a Natural Resources Mining Client and developed a Predictive Model which provided recommendations based on the buying patterns using Association Rule.

• Fraud Detection: Built classification model to detect if the claim filed for an auto insurance company is a fraud or not with an accuracy of 78%.

• Promoted to the role of Product Risk Analyst to handle data of Mark to Market Risk models. Analyzed and performed financial analysis for new profit margins /loss and recommend customer on the ways to improve it.

• Awarded as the Fresher of the year for contributing towards analytical skills, Business Operations Research, Presentation, Communication skills Engage, Mumbai, India Feb 2011 – Oct 2013

Research Analyst

• Designed Use Cases and fetched data using complex SQL Queries from student’s data from various colleges.

• Developed a Customer Targeting tool to target specific people with phone calls and emails for fund raising and sending invitations for upcoming events.

• Designed interactive dashboard using Tableau to visualize campaign and event data for getting better insights. ANALYTICS PROJECTS

• Aviage Systems: Predicted fuel burn rate of an aircraft using SVR and XGBoost in R and recommended speed change to minimize fuel burn rate. Achieved a prediction accuracy with an impact of 92.3% which saved fuel by 6.7%.

• IBM Watson: Created more than 10 dashboards to visualize HR data on IBM Watson platform and gave meaningful insights and recommendations.

• Built Clustering models using K-means and Kohonen using SPSS to characterize the consumption pattern based on annual spending.

• Santander Bank: Identified satisfied or unsatisfied customers at start of service. Implemented two class boosted decision tree and two class decision jungle with Area under curve of 81.5% and Accuracy of 95.1%.

• Telstra Networks: Implemented models with Multiclass algorithms of Neural Network, Logistic Regression, Decision Jungle with performance criteria of Accuracy, AUC. Merged multiple data frames using dcast function in R. Validated using 7-fold cross validation with tuning to increase the accuracy of the training set. Achieved an accuracy of 88% using Stacked Ensemble with majority voting. Ranked in top 20% on Kaggle board with a score of 0.54.

• Amazon Meta Data: Built recommendation system with item and user based collaborative filtering in Python based on different similarity measures.

• House Price using Regression Techniques: Executed mice technique to impute missing data in R. Built regression models based on Bayesian Linear Regression, Poisson Regression, Logistics Regression, with the accuracy rate of predicting house rate for consumers in advance up to 85%.

• Kolby’s Corner: Maximized profit margin for Kolby’s joint using Constant Demand Elasticity function which increased sales up to 21.7%. CERTIFICATIONS

• Google Analytics: Analytics for Beginners, Advanced Analytics, E-Commerce and applied Analytics

• Johns Hopkins University.: Data Scientist’s Toolbox, R Programming, Getting & Cleaning Data, Statistical Inference, Practical Machine Learning, EDA

• British Computer Society: Certificate in Business Analysis and ITIL R Foundation: ITIL for continuous service improvement

Contact this candidate